1 paper · 1 filter
Yuanhao Zhai, Ziyi Liu, Zhenyu Wu +5
Deep learning models have a risk of utilizing spurious clues to make predictions, such as recognizing actions based on the background scene. This issue can severely degrade the ope…